BINARY STAR ORBITS. II. PRELIMINARY FIRST ORBITS FOR 117 SYSTEMS Diana M. Seymour, 1 Brian D. Mason, William I. Hartkopf, and Gary L.

Transcription

1 Te Astronomical Journal, 123: , 2002 February # Te American Astronomical Society. All rigts reserved. Printed in U.S.A. BINARY STAR ORBITS. II. PRELIMINARY FIRST ORBITS FOR 117 SYSTEMS Diana M. Seymour, 1 Brian D. Mason, William I. Hartkopf, and Gary L. Wycoff US Naval Observatory, 3450 Massacusetts Avenue, NW, Wasington, DC ; dseymour@fas.arvard.edu, bdm@draco.usno.navy.mil, wi@usno.navy.mil, glw@draco.usno.navy.mil Received 2001 September 26; accepted 2001 October 29 ABSTRACT Orbital elements are presented for 117 binary systems wit no previous orbit calculation. For nearly all of tese systems, tese elements must be regarded as preliminary, but te epemerides presented ere sould be relatively accurate over te next several decades. Furter, te analysis of tese systems sould igligt te need for teir continued observation by dedicated programs to improve te veracity of tese elements. Key words: binaries: general binaries: visual 1. INTRODUCTION As double star observing programs continue, investigation of systems wose motion years ago was indeterminate can now reveal Keplerian motion. A project was undertaken to examine systems in te Wasington Double Star Catalog (WDS; Mason et al. 2001) and pinpoint tose tat, wile aving no publised orbits, now sow orbital motion. For 117 of tese systems, te observational data were sufficient to calculate first orbits. Tese orbits are not expected to be definitive; rater, tey are meant to provide a baseline for future observations and orbit calculations. As most of tese systems ave too few data to be well defined, te epemerides are expected to be more useful tan te orbital elements in most cases. We calculate tese orbits in part to encourage furter observation of tese systems, many of wic ave gone unobserved for a decade. Te systems considered as candidates for tis project met tree criteria. First, tey did not ave publised orbits. Second, tey needed a minimum of eigt observations so as to avoid making broad pronouncements based on little data. Finally, tey ad to exibit at least a 30 cange in position angle or a 30% cange in separation. Te observations of eac candidate system were plotted so tat eac system could be visually examined. Some systems, wic will be te subject of future analysis, displayed rectilinear motion, probably due to differential proper motion; oters sowed uncertain motion; still oters appeared to be stationary. In te latter case, incorrect quadrant assignments or observational errors were responsible for teir selection. Orbits were attempted only for tose systems tat visually displayed significant curvature. Initial estimates of orbital period, epoc of periastron passage, and eccentricity (P, T 0,ande) are made, based on te motion indicated by te available data, along wit grid step sizes for tose elements. Weigts of observations are assigned on te basis of telescope aperture and observing tecnique, as described in Hartkopf, Mason, & Worley (2001). Orbits are ten calculated according to te treedimensional adaptive grid searc tecnique, initially described by Hartkopf, McAlister, & Franz (1989) and modified as described by Mason, Douglass, & Hartkopf 1 Science and Engineering Apprenticesip Program intern (1999). Briefly, te four Tiele-Innes elements (A, F, B, and G) are calculated by te metod of least squares. From A, F, B, and G te oter geometric elements (a, i,, and!) are calculated directly. As tis procedure is performed, te initial values of P, T, and e are modified by an adaptive grid searc. Over several iterations, observations are assigned new weigts according to teir rms errors and residuals, and te orbital elements eventually converge as te grid searc narrows. 2. NEW ORBITS Of te more tan 84,000 systems in te WDS, 2302 systems wit 54,718 observations satisfied te criteria enumerated in te previous section and were considered candidate systems. Of tese, 255 initially appeared to display Keplerian motion. Closer examination of te data, before, during, and after attempts at calculating orbits, led to a final result of 117 serviceable orbits incorporating 3356 observations (see Figs. 1 10). In sifting troug tese data, several undred erroneous measures were discovered. Among tose systems wit orbits publised ere, 213 measures wit flipped quadrants were found; tese were corrected for te orbit calculations by adding or subtracting 180 from eac measure as appropriate. An additional 422 measures wit flipped quadrants were gleaned from systems wose orbits could not be calculated at tis time. Werever possible, tese quadrant ambiguities were resolved using differential magnitude measurements. For most systems, were only a measure or two ad quadrants inconsistent wit te remainder, only tese outliers were canged. Furter, 15 measures reported angular separations so vastly different from oter observations of te same system tat tey were removed entirely from system plots (for considerations of scaling) and were never incorporated into orbit calculations. Two of tese belonged to systems wit orbits publised ere (HU 1213 and HU 190). Tese measures ave been flagged in te WDS as possibly erroneous. Te results of our orbit calculations are presented in Table 1. Te first column sows te epoc J coordinates of eac system, wit rigt ascension given to te nearest tent of a minute of time and declination given to te nearest minute of arc. Te second column gives te discoverer designation, wit components specified if any additional ones are known to exist. Te tird troug nint

2 Fig. 1. Orbital information for (left to rigt and top to bottom) = A 1256 AB, = D 2 AB, = HU 1012, = A 651, = A 924, = B 644, = COU 1654, = LDS 873, = MLR 87, = BU 235 Aa, = I 264 AB, and = A 948 AB. In tis and subsequent figures, orbits are plotted as te pat of te secondary star around te primary. Eac observation ( plus signs, visual measurements; asterisks, potograpic measures; H, Hipparcos measures; T, Tyco measures; filled circles, speckle interferometry; open circles, optical interferometry) is connected to its predicted position by an O C line, wic may be interpreted as an error bar; a dotted O C line signifies a zero-weigted measure. Te line of nodes is te das-dotted line passing troug te primary star. Te scale is in arcseconds. 1024

3 Fig. 2. Same as Fig. 1, but for = A 2602, = HU 1213 (one zero-weigted measure as been omitted for purposes of scaling), = I 265, = HU 16, = A 450, = HU 1350, = HU 539, = SEE 23, = A 1289, = A 831, = STF 567, and = RST

4 Fig. 3. Same as Fig. 1, but for = A 481, = DON 97, = RST 2375, = RST 4800, = A 670, = A 1061 AB, = A 2122, = WRH 15 AB, = HJ 4087 AB, = B 1115, = B 1122, and = I

5 Fig. 4. Same as Fig. 1, but for = I 843 AB, = HU 1594, = B 194, = STT 218, = STT 220, = I 212, = I 506, = A 2575, = RST 4944, = I 885, = A 7, and = STT

6 Fig. 5. Same as Fig. 1, but for = HU 730, = RST 3767 AB, = A 1088, = A 1999, = A 78, = CPO 12 A-BC, = WRH 12, = HU 739, = RST 1712, = I 298, = RST 3844, and = ES 608 AB. 1028

7 Fig. 6. Same as Fig. 1, but for = BU 343, = SWI 1, = I 1243 AB, = RST 2917, = A 1109 AB, = STF 1876 AB, = A 2172, = B 1288 AB, = BU 947 AB, = MCA 42 CE, = A 1798, and = STT

8 Fig. 7. Same as Fig. 1, but for = VBS 26 AB, = MLR 198, = BU 129, = HU 177, = HU 751, = B 915 AB, = HU 1285, = HU 1182, = HU 190 (one zero-weigted measure as been omitted for purposes of scaling), = A 1374, = HU 674, and = HO

9 Fig. 8. Same as Fig. 1, but for = B 1352, = HU 1291, = HU 66 AB, = B 2546 Aa, = RST 2073, = A 2989, = I 1390, = A 713, = A 1400, = BU 146, = B 459 BC, and = MLR

10 Fig. 9. Same as Fig. 1, but for = BU 763 AB, = HU 1615, = A 613, = B 1001, = BU 678 AB, = HU 1626, = A 2289 AB, = HU 372, = HU 374, = LV 10, = I 1449, and = STF 2872 BC.

11 BINARY STAR ORBITS. II Fig. 10. Same as Fig. 1, but for = WOR 11, = HU 389, = HU 985, = DON 1038, = RST 3320, = HEI 88, = A 1494, = A 1247, and = BU 730. columns give orbital elements, wic are reported wit a precision appropriate to teir errors. Te tird column gives te orbital period (P) in years; te fourt gives te semimajor axis in arcseconds (a 00 ); te fift, inclination (i) in degrees; te sixt, te longitude of nodes () in degrees; te sevent, time of periastron (T 0 ) expressed as a fractional Besselian year; te eigt, eccentricity (e); and te nint, te longitude of periastron (!) in degrees. Te 10t column gives te orbit grade, for wic a few words of explanation are needed. In te Fourt Catalog of Orbits of Visual Binary Stars, Worley & Heintz (1983) assigned grades between 1 ( definitive ) and 5 ( indeterminate ) to te publised orbits. Grades in te 10t column of Table 1 were determined by te quantitative metod described in Hartkopf et al. (2001). Tis metod approximates objectively te subjective judgments of Worley & Heintz, wo, as two of te tree most prolific double star observers of all time, ad personal experience wit stars, equipment, and observers tat is now impossible to duplicate. Tis quantitative grading sceme is based on, among oter tings, te and pase coverage of te observations, number of revolutions traversed between te earliest and most recent observations, number of observations, and weigted rms residuals. Generally, te lower te rms residuals and te greater te coverage, te better te grade of te orbit will be. Most of te orbits publised ere received a grade of 4 ( preliminary ) or 5 ( indeterminate ) due to te lack of coverage; only te best eigt were deemed to be of grade 3 ( reliable ). Epemerides for te period are presented in Table 2, te first two columns of wic are identical to tose

12 TABLE 1 Orbital Elements WDS Coordinate, (J ) Discoverer Designation P (yr) a i T (yr) e! Grade A 1256 AB D 2 AB HU A A B COU LDS MLR BU 235 Aa I 264 AB A 948 AB A HU I HU A HU HU SEE A A STF RST A DON RST RST A A 1061 AB A WRH 15 AB HJ 4087 AB B B I I 843 AB HU B STT STT I I A RST I A STT HU RST 3767 AB A A A CPO 12 A-BC WRH HU RST I RST ES 608 AB BU SWI I 1243 AB RST

13 BINARY STAR ORBITS. II TABLE 1 Continued WDS Coordinate, (J ) Discoverer Designation P (yr) a i T (yr) e! Grade A 1109 AB STF 1876 AB A B 1288 AB BU 947 AB MCA 42 CE A STT VBS 26 AB MLR BU HU HU B 915 AB HU HU HU A HU HO B HU HU 66 AB B 2546 Aa RST A I A A BU B 459 BC MLR BU 763 AB HU A B BU 678 AB HU A 2289 AB HU HU LV I STF 2872 BC WOR HU HU DON RST HEI A A BU in Table 1. Te subsequent 10 columns give te predicted position angle (, measured in degrees and corrected for precession) and angular separation (, measured in arcseconds) for eac of te epocs , , , , and As tese orbits are preliminary only, te epemerides given in Table 2 are expected to be bot more accurate and more useful tan te elements given in Table NOTES ON INDIVIDUAL SYSTEMS = I 265: Our finding of as te time of periastron supports te prediction ( a close approac in te 1970 s ) made in te WDS = B 194: Altoug Hartkopf et al. (1993) predicted a time of periastron around , we find te

14 TABLE 2 Epemerides WDS Coordinate, (J ) Discoverer Designation A 1256 AB D 2 AB HU A A B COU LDS MLR BU 235 Aa I 264 AB A 948 AB A HU I HU A HU HU SEE A A STF RST A DON RST RST A A 1061 AB A WRH 15 AB HJ 4087 AB B B I I 843 AB HU B STT STT I I A RST I A STT HU RST 3767 AB A A A CPO 12 A-BC WRH HU RST I RST ES 608 AB BU SWI

15 BINARY STAR ORBITS. II TABLE 2 Continued WDS Coordinate, (J ) Discoverer Designation I 1243 AB RST A 1109 AB STF 1876 AB A B 1288 AB BU 947 AB MCA 42 CE A STT VBS 26 AB MLR BU HU HU B 915 AB HU HU HU A HU HO B HU HU 66 AB B 2546 Aa RST A I A A BU B 459 BC MLR BU 763 AB HU A B BU 678 AB HU A 289 AB HU HU LV I STF 2872 BC WOR HU HU DON RST HEI A A BU time of periastron to be Te discrepancy may be explained by tree measures not publised at te time of te 1993 paper = BU 947 AB and MCA 42 CE: Orbits for two component pairs of tis complex star system were calculated independently of one anoter. A complex, multibody solution was not attempted = HU 66 AB: Van Biesbroeck (1954) was unable to resolve tis system in 1943, 1944, or 1945 on te 82 inc (2.1 m) telescope at te McDonald Observatory.

16 1038 SEYMOUR ET AL. However, since e resolved te system successfully in 1946 wit te same telescope, and in 1940 wit a smaller (40 inc) telescope, te nonresolution is probably due to bad seeing or identification error. 4. CONCLUSIONS Wile te orbits presented in Table 1 can ardly be considered reliable over te course of even a single revolution, te epemerides in Table 2 sould be quite adequate for te next decade. Indeed, it migt be said tat te greatest use of te elements from Table 1 is for te prediction of and for epocs not given in Table 2. It is oped tat inclusion of tese objects in dedicated observational programs may ascertain te veracity of tese orbital elements. Tanks are provided to te US Naval Observatory for teir continued support of te double star program, and to te Science and Engineering Apprenticesip Program (SEAP-CQL), a joint effort by te Department of Defense and George Wasington University, wic also sponsored tis researc. Tis researc as made use of te SIMBAD database, maintained and operated at CDS, Strasbourg, France. Hartkopf, W. I., Mason, B. D., Barry, D. J., McAlister, H. A., Bagnuolo, W. G., & Prieto, C. M. 1993, AJ, 106, 352 Hartkopf, W. I., Mason, B. D., & Worley, C. E. 2001, AJ, 122, 3472 Hartkopf, W. I., McAlister, H. A., & Franz, O. G. 1989, AJ, 98, 1014 Mason, B. D., Douglass, G. G., & Hartkopf, W. I. 1999, AJ, 117, 1023 Mason, B. D., Wycoff, G. L., Hartkopf, W. I., Douglass, G. G., & Worley, C. E. 2001, AJ, 122, 3466 REFERENCES van Biesbroeck, G. 1954, Measurements of Double Stars (Publ. Yerkes Obs., 8, 159) (Cicago: Univ. Cicago Press) Worley, C. E., & Heintz, W. D. 1983, Fourt Catalog of Orbits of Visual Binary Stars, (Publ. US Nav. Obs., 2d Ser., 24, Part 7) (Wasington: GPO)